My first experience undertaking a cost-of-illness study dates back to the early 1990s. To set the context, President Clinton had just introduced his Administration’s healthcare initiative. Just a few years earlier, Eli Lilly had come out with the blockbuster antidepressant, Prozac®, the first selective serotonin reuptake inhibitor (SSRI) (with Zoloft®, Paxil®, Celexa®, Lexapro®, and Luvox® eventually to follow). Soon, much would change in terms of recognition and treatment of psychiatric disorders.

In those days, depression—like most mental illnesses—was stigmatized and vastly undertreated. So, in 1992, Eli Lilly came up with an interesting idea. As the market leader in the SSRI antidepression space, it decided to focus on promoting not only its particular treatment, but also awareness of the underlying disease.

Eli Lilly approached our research team, and funded a cost-of-illness study. Tipper Gore, wife of the Vice President and designated point person on the Clinton Administration’s mental health initiative, became the public spokesperson for the study when it was released at the National Press Club in Washington, DC in 1993 [1]. Through this experience, I came to understand the immense value of publicizing this work. Looking back, it undoubtedly contributed to a national conversation that positively affected social attitudes and norms.

The press conference was very widely attended and our results were very well-received. A great many news stories followed in print, radio, and television media outlets around the country describing in detail how depression is highly prevalent; that it is not so much a disease of aging, but rather tends to affect people from an early age of onset; that it is not so debilitating that it requires most sufferers to withdraw entirely from the labor force, but that it affects employers in terms of both absenteeism and presenteeism in every workplace; and that the direct costs of illness were just a small piece of what we estimated to be a substantial overall economic burden. That cost-of-illness article ended up being very highly cited, and was one strand in the ensuing discussion about de-stigmatizing a highly prevalent and pernicious mental illness. In the years since, my colleagues and I have built on those early findings with many more cost-of-illness studies on other aspects of depression and its treatment.

I thought about that vignette while reading Steven Kymes’s article “Can we declare victory and move on?” The case against funding burden-of-disease studies [1]. The article makes a very useful point: cost-effectiveness analysis provides policy makers a needed tool with which to allocate health care resources. He notes that by focusing on costs alone, burden-of-illness studies are of only limited value in this context. Therefore, he concludes that we in the research community ought to steer funders away from the latter and towards the former given the limited funds available for pharmacoeconomic investigations.

But Dr. Kymes’s point is premised on the idea that all pharmacoeconomic research is intended to directly affect payer decision making. As my own personal anecdote highlights with respect to the economic burden of depression, that may not always be the primary objective of this line of research, even if it is invariably an eventual objective. There are other stakeholders in the context of payer decision making, such as patients, prescribers, and Congressional leaders, for whom introducing the calculation of cost per quality-adjusted life-year could have the unintended consequence of shutting down rather than spurring continued discussion. I suspect that consideration of cost of illness (which focuses only on costs) or comparative effectiveness research (which focuses only on effectiveness) would, at times, be a more inviting point of entrée into important resource allocation discussions than would the results of a traditional cost-effectiveness analysis. Looking forward, I see many opportunities for cost-of-illness research findings to continue to inform such conversations about healthcare resource utilization.